Detecting and recognizing a color of yellow lane markers on a road is important to autonomous driving. The yellow lane markers are conventionally used to separate traffic flow in opposite directions and to separate the road from pavement edges. The color recognition problem is challenging when the lane markers are faded, degraded and when lighting conditions are poor. Furthermore, conventional color recognition techniques are computationally demanding and thus difficult to perform in real time.
It would be desirable to implement a real-time color classification for street vehicles.